Prediction of BMS number by image analysis and comparison of the feature quantity of marbling in different cross sections of Holstein steers
2005
Hamasaki, Y.(Obihiro Univ. of Agriculture and Veterinary Medicine, Hokkaido (Japan)) | Kuchida, K. | Hidaka,S. | Shimada, K. | Sekikawa, M. | Naruyama, S.
The 6-7th rib of carcass cross section has been used in meat quality evaluation of beef in Japan. Marbling scores may have some discrepancy in different sections even for the same muscle. The aims of this study were to develop a prediction method of the BMS (Beef Marbling Standard) number by image analysis and to investigate the difference of the feature quantity of marbling in the 6-7th ribs and those in other cross sections. Digital images of the 6-7th cross section from 61 Holstein steers were used to predict the BMS number with a multiple regression equation. Additionally, digital images of the 4 cross sections, which were cut in 2.5 cm intervals from the 6-7th cross section toward the lumbar, from 18 other Holstein steers were used to compare of the feature quantity of marbling in different cross sections. The ratio of marbling to rib eye area (FATPER), the coarseness of marbling and the shape of rib eye were considered as independent variables, and the BMS number evaluated by a grader was a dependent variable for the multiple regression analysis. Selected variables of the multiple regression equation for estimating the BMS number were FATPER, the area of the largest marbling particle, coarseness of a single marbling particle and rib eye area (R2=0.71). The most contributing variable in the prediction of BMS number was FATPER, and the partial regression coefficient was 0.54. From investigating the FATPER and the prediction BMS number in the 4 cross sections, one case was found in which there was a large difference in FATPER between adjacent cross sections (15.9%-26.2%). However, there was only a slight difference in the prediction BMS numbers among the 4 cross sections (1.67-2.17) in this case.
اظهر المزيد [+] اقل [-]